Modern organizations and corporations often rely on computer systems to manage the massive amounts of data acquired in the course of running their operations. Interpreting large volumes of data often presents great challenges to those responsible for understanding the data. A common method for visualizing large volumes of data is to use a bar chart. A bar chart is a type of graph in which different values are represented by rectangular bars.
In a typical bar chart, each bar represents a single data value. It is common to represent aggregated data using two-dimensional or three-dimensional bar charts. When using a bar chart, the number of data items shown at the same time must be pre-decided. Additionally, the bar chart can only show a small range of data items (10–20), because showing too many items will obscure the data. Furthermore, if a three-dimensional bar chart is used, some bars hide others and overlapping makes the visual impression misleading, as the area seen is not proportional to the value to be represented. In these cases, valuable information often gets lost.
FIG. 1 illustrates a three-dimensional bar chart 100 of over 600,000 customer orders and corresponding purchase analysis. To allow for viewing, all data is aggregated by order 110, price 120 and quantity 130. No individual customer information is shown, only the aggregate total of customer information. The display is difficult to understand, due to some data (e.g., column 140) being disproportionately larger than other columns, making it difficult to understand the values of the smaller columns. Furthermore, some columns (e.g., column 140) hide other columns, making that data extremely difficult, if at all possible, to understand.
Another common method used to visualize large volumes of data is a stack chart 200 as shown in FIG. 2. A stack chart is similar to a bar chart, but also allows for the area of a color to be proportional to the value of a particular set of data.
Stack chart 200 illustrates an exemplary stack chart illustrating sources of energy in the United States by year. The area of each color is proportional to the amount of energy provided by the respective source in that year. Each year is a group. Only five groups are shown in FIG. 2. Limiting the number of groups to five allows for easy viewing of the data for the selected years, but severely limits the amount of data displayed. It may be desirable to view more data than just the five years displayed. Increasing the number of groups, for example by showing all 372 months from 1950 until 1980, results in an overwhelming amount of data that is difficult to understand and interpret. If stack chart 200 were displayed on a standard computer system display of 1024×768 pixels, the groups would have a width of three pixels. These narrow groups do not allow the showing of small width variations accurately like those from 1970 until 1980 (e.g., 2.4, 3.1, 2.7) in stack chart 200. In addition, at least 25% of the screen space is wasted for gaps between the groups, even if only one pixel is used as a gap.
Due to constraints dictated by the viewing limits of human eye, both bar charts and stack charts are required to reduce the number of values shown as the number of records increases. Otherwise, the chart obscures the nature of the data, as the chart will contain too much data in too small a space for a human reader to process and understand. The number of values can be reduced by limiting the number of groups showing or by aggregating data. Both selection and aggregation reduce the information seen, limiting access to more detailed data.
Accordingly, a need exists for a method for visualizing large volumes of data having multiple attributes without requiring aggregation of the data. A need also exists for a method that accomplishes the above need and allows for a better way to compare records and to identify trends and patterns in data. Additionally, a need exists for a method that accomplishes the above needs and allows for direct access to the detail data by drilling down at single data items. Furthermore, a need exists that accomplishes the above needs and is easily understood by a user.